Home › Work

Case studies

Software we have designed, built and run for clients in the United States. Client names are withheld under agreement and shared on request; figures are the client’s own.

Home improvement e-commerce · National US online retailer

An AI cabinet visualiser that lifted leads 20% for a national online retailer

A generative AI visualiser, built into an existing nopCommerce storefront, lets shoppers see a cabinet door style in their own kitchen before they order. Leads rose 20% and conversion to orders 5%.

Home improvement e-commerce · the same national retailer

Sales forecasting that knows the calendar, the map and the weather

Years of order history in a nopCommerce database became a forecasting model that predicts demand day by day and month by month, by region, with separate handling for sale peaks and regional climate seasons.

Industrial manufacturing · US weighing-equipment maker

A B2B catalogue and AI knowledge base that halved customer-service answer times

Engineers buying weighing equipment now find parts by application, request quotes and pull certificates themselves, and support answers what remains from a 900-article library in half the time.

Wholesale distribution · supplier to a national carrier’s retailers

An ERP-connected store that stopped finance re-keying payments

A distributor’s ageing store did not talk to its ERP, so inventory, orders and payments were reconciled by hand. The rebuilt store keeps 1,100+ SKUs, orders and payments in step with NetSuite.

Real estate · US coastal brokerage

An AI assistant that answers agents from the brokerage’s own knowledge

Agents at a coastal brokerage asked the same questions of an outdated knowledge base. A Bedrock-powered assistant, with Elasticsearch as the retrieval layer, answers from company resources inside the intranet they already use.

Mobility · US client, in progress

A real-time ride-hailing platform for customers, drivers and administrators

A new ride-hailing service needed a backend that matches drivers to riders in seconds and stays consistent while many bookings change at once. The platform is built and documented, and is preparing for launch.